This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCC28730: Unexpected Switching

Part Number: UCC28730
Other Parts Discussed in Thread: PMP9009, ,

Hi All,

My customer is designing an isolated power supply independently from battery for a forklift application.

I was asked to find a demoboard with below specs for the power part:


  • Vin Min <= 24V; V max >= 79 V (DC)
  • VinNom 48V
  • Vout= 12V, isolated
  • Absorbs the minimum current possible with no load even though enabled (Low Power)
  • IoutMax>=2A, Iout Nom 0,5A.
  • No optocoupler

I was already in contact thru email with PDS team and they suggested me PMP9009, that's a good design but does not meet all above specs.

Apart from a custom design could you please support me for the below technical questions?

Customer while using the UCC28730 demo board found some problems not solved in the ds of UCC28730, which states that the controller imposes a maximum switching frequency of 83kHz instead from the measurements made by customer it emerges that at the start  between the second and third commutation there is a period ranging from 10us to 6.4us against the 12us estimated.

This means that the diode at the secondary has a conduction time such that its duty cycle exceeds 0.432 (DMAGCC).

Customer also found out that this period of time changes by disconnecting the C6, which connects the masses between primary and secondary, and by shorting it or using it normally in none of the 3 cases the device starts.

Could you please help me in understanding this unexpected switching?

I am attaching the screenshot of what said above (screenshot1).

I also attach the images of the tests customer carried out on the component, with a brief description:

 All tests were carried out with a supply voltage of 28 to 40V.

Description of the images

1- In the IMG ... WA0007 the voltage on the TP2 is shown in blue, feeding the ucc28730. And in purple the output voltage. It clearly shows that the device enters protection

2- In the VERDECS image ... It is shown with the green trace: CS TP5, Viola Vout +; Vaux blue; yellow Vdrain of q1.

3- In IMG ... WA00022 it is shown that the switching frequency exceeds 83kHz, also there was a slow diode which I replaced with a schotcky which will find the pN in the schematic.

4- In IMG ... WA00030 you can see how the oscillation is sinusoidal having changed the diode on the secondary.

5- In IMG… 095451 you will find a photo of the schematic the transformer used is 7491195212 with an L of 21uH and a turns ratio between Nps = 1.33 / 1 and npa = 1.33 / 0.89

The installed shunt has a value of 0.3 Ohm contrary to what is reported to prevent the secondary diode from exceeding the DMAGCC.

Any further comments?

  • Hello Antonio,


    Thank you for your interest in the UCC28730 flyback controller.


    Although the PMP9009 reference design does not meet all of the customer specs, it is still useful as a reference of how the design will look. Its parameters can be adjusted so that a similar design does meet the customer specs.


    The higher frequency and longer DMAGCC time during start-up is explained in Section 7.3.7, page 19, of the UCC287830 datasheet. Start-up is a special operating case, and this behavior is normal.


    I am not clear on the issue with removing C6. On the UCC28730EVM-552 evaluation board, that cap is a common-mode EMI by-pass capacitor to reduce conducted EMI. Removing it should have no effect on start-up. Shorting it out should also have no effect PROVIDEDED that the AC source to the EVM is isolated from Earth GND.


    You mention an attachment of images, but there is no file attached, so I can’t really comment on them.

    But in general, there is no need to attempt to force DMAGCC to = 0.432, nor force the switching frequency to run < 83kHz. The IC will set these parameters according to its operating condition, during start-up and after start-up.